如何制作一个Node命令行图像识别工具

# 创建 recognition 项目
mkdir recognition
cd recognition
npm init -y

# 安装主依赖
yarn add images tesseract.js

# 安装工具依赖
yarn add chalk yargs

# 可选依赖
yarn add socks5-http-client

#!/usr/bin/env node

// package.json
{
 "bin": {
  "reg": "./cli.js"
 }
}

reg --url=https://static.chenng.cn/imgs/test_img.png

yarn link

const Tesseract = require('tesseract.js');
const images = require('images');
const requset = require('request');
const fs = require('fs');
const { promisify } = require('util');
const chalk = require('chalk');

const writeFile = promisify(fs.writeFile);
const rp = promisify(requset);

class Recognize {
 constructor(url) {
  Recognize.downloadDir = `${__dirname}/dist/`;
  Recognize.downloadFile = `${__dirname}/dist/temp.png`;
  this.url = url;
  this.start();
 }

 async start() {
  const data = await this.downloadImg();
  await writeFile(Recognize.downloadFile, data);
  this.recognize();
  const result = await Tesseract.recognize(Recognize.downloadFile, {
   lang: 'eng',
   tessedit_char_blacklist: 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ',
  });

  console.log(`

   识别成功!
   识别结果为：${chalk.green(result.text)}
  `);
 }

 async downloadImg() {
  if (!fs.existsSync(Recognize.downloadDir)) {
   fs.mkdirSync(Recognize.downloadDir);
   console.log(`创建了 ${Recognize.downloadDir} 文件夹`);
  }
  
  const res = await rp({
   url: this.url,
   method: 'GET',
   encoding: null,
  });
  return res.body;
 }

 recognize() {
  // 放大图片，并覆盖源文件
  images(Recognize.downloadFile)
   .size(400)
   .save(Recognize.downloadFile);
 }
 
}

module.exports = Recognize;

# 新建代码仓库，git push

# 登录到 npm
npm adduser

# 发包
npm publish --access public

# 全局安装
npm install -g @chenng/recognition